madlag's picture
Initial version of 75% sparsity BERT finetuned on SQuAD1.1.
c8c970e
{
"_name_or_path": "madlag/bert-base-uncased-squad-v1-sparse0.25",
"ampere_mask_init": "constant",
"ampere_mask_scale": 0.0,
"ampere_pruning_method": "disabled",
"architectures": [
"BertForQuestionAnswering"
],
"attention_probs_dropout_prob": 0.1,
"gradient_checkpointing": false,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"hidden_size": 768,
"in_shuffling_group": 4,
"initializer_range": 0.02,
"intermediate_size": 3072,
"layer_norm_eps": 1e-12,
"mask_block_cols": 32,
"mask_block_rows": 32,
"mask_init": "constant",
"mask_scale": 0.0,
"max_position_embeddings": 512,
"model_type": "bert",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"out_shuffling_group": 4,
"pad_token_id": 0,
"pruning_method": "topK",
"pruning_submethod": "default",
"shuffling_method": "disabled",
"type_vocab_size": 2,
"vocab_size": 30522
}